Cognitive Errors in Diagnostic Reasoning

Transcription

1 A9 Cognitive Errors in Diagnostic Reasoning Tami Wallace, DNP, APRN, NNP-BC Neonatal Nurse Practitioner Monroe Carell Jr. Children s Hospital at Vanderbilt Nashville, TN The speaker has signed a disclosure form and indicated she has no significant financial interest or relationship with the companies or the manufacturer(s) of any commercial product and/or service that will be discussed as part of this presentation. Session Summary Every medical professional would like to be perfect; none of us would choose to make errors. Unfortunately, we do make errors. The human brain is designed to look for patterns, make conclusions based on pattern recognition, and has built in biases that are used in decision making. This presentation will examine why we make errors, the types of errors we make, how those errors change with experience, and what we can do to try to prevent them. Session Objectives Upon completion of this presentation, the participant will be able to: discuss incidence and recognition of cognitive and diagnostic errors; understand the effect of various individual factors on the incidence and type of errors; implement strategies to prevent cognitive and diagnostic errors. References Berk, et al. (2008). The effect of clinical experience on the error rate of emergency physicians. Annals of Emergency Medicine, 52(5): Berner & Graber (2008). Overconfidence as a cause of diagnostic error in medicine. American Journal of Medicine, 121 (Suppl 5): S2-S23. Brodlie, Laing, Keeling & McKenzie (2002). Ten years of neonatal autopsies in tertiary referral centre: Retrospective study. BMJ, (324): Croskerry (2003). The importance of cognitive errors in diagnosis and strategies to minimize them. Academic Medicine, (78): Croskerry (2009). Clinical cognition and diagnostic error: Applications of a dual process model of reasoning. Advances in Health Sciences Education, (14): Croskerry (2009). A universal model of diagnostic reasoning. Academic Medicine, 84(8): Croskerry (2009). Context is everything or how could I have been that stupid? Healthcare Quarterly, 12: p e171-e177.

2 Croskerry, Singhal, Mamede (2013). Cognitive debiasing 1: Origins of bias and theory of debiasing. BMJ Quality and Safety, 22: ii58-ii64. Elder & Zuccollo (2005). Autopsy after death due to extreme prematurity. Archives of Disease in Childhood, Fetal & Neonatal Edition, 90(3): Ericsson (2008). Deliberate practice and acquisition of expert performance: A general overview. Academic Emergency Medicine, 15(11): Eva (2002). The aging physician: Changes in cognitive processing and their impact on medical practice. Academic Medicine, 77(10): s1-s6. Graber (2009). Educational strategies to reduce diagnostic error: Can you teach this stuff? Advances in Health Sciences Education, 14: Harasym, Tsai & Hemmati (2008). Current trends in developing medical students critical thinking abilities. Koahsiung Journal of Medical Sciences, 24(7): Norman & Eva (2010). Diagnostic error and clinical reasoning. Medical Education, 44: Potchen, et al. (2000). Measuring performance in chest radiography. Radiology, 217(2): Singh, Peterson & Thomas (2006). Understanding diagnostic errors in medicine: A lesson from aviation. Qualified Safety in Health Care, 15: Session Outline See presentation handout on the following pages.

3 Tami Wallace DNP APRN NNP BC I don t make cognitive errors, Do you? Diagnostic errors. Cognitive errors and frequent types of errors. Does experience change the incidence or types of errors? Do Nurse Practitioners make different types of errors? Strategies to prevent cognitive errors. Can you remember making any errors in diagnosis? In this era of safety, why aren t we talking about these? ½ of all litigation against emergency physicians is due to delayed or missed diagnosis Emergency medicine has been called the natural laboratory of error (Croskerry, 2003) Reading xrays requires perception and cognition 100 radiologists, shown 60 xrays and asked Is the film normal Disagreed between themselves 20% of time When same radiologists reread films.contradicted their own analysis 5 10% of time One film missing clavicle (requires noticing what is not there). 60% failed to identify.if asked to look for cancer 83% found it. Potchen,et al 2000

4 Residents and all other subspecialties were more inconsistent reading xrays than the radiologists.. How many films have you looked at this week? Very little information available about our diagnostic accuracy We haven t looked Literature search on autopsies in neonates and concordance with neonatal diagnosis..eliminated studies from third world countries ti Autopsy reports: Two studies remained 29 autopsies on very preterm infants Very preterm infants less likely to have an autopsy New findings in 79%, significant change in diagnosis 28% Elder, years of autopsy reports 74% had complete concordance between clinical cause of death and autopsy. New information in 26% Brodlie, 2002 There are Three types. No Fault errors System errors Cognitive errors (Harasym, 2008) (Croskerry, 2003) The disease is present, but not detected. Disease is silent, presents atypically or mimics something more common Case: Not all hoof beats are horses Sometimes you have to hunt for Zebras

5 Diagnosis is delayed or missed because of imperfections in the health care system. Case 1: the official radiology report of xrays is dictated after clinicians have seen films and reacted to them. The reports are not routinely reviewed. Misdiagnosis from faulty data collection or interpretation, flawed reasoning, or incomplete knowledge. Limitations in processing, using heuristics (short cuts/ pattern recognition) Memory limitations and excessive cognitive loading Bias Cognitive Errors The majority of diagnostic failures (75%) are due to cognitive error (Graber, 2005) Cognitive dispositions to respond (CDRs) Failure in perception Failed heuristics Biases Croskerry, 2003 Decision Making Intuitive Type 1 Dual process theory Type 1 Type 2 Analytical Type 2 Inductive, heuristic, reflexive, skilled, rules of thumb, pattern recognition Advantages: Fast!!!!, minimal effort Low cost Disadvantages: Vulnerable to bias (some may be hardwired) Errors common Highly dependent on context Low dependence on scientific rigor Analytical, normative, deductive, abstract Advantages: deliberate and rule based, less vulnerable to bias, few errors, context less important Disadvantages: slow, deliberate, high cost, learned (not hardwired)

6 Your brain looks for patterns It then tends to ignore information that does not fit the pattern Your brain has a limited ability for attention Finished files are the result of years of scientific study combined with the experience of years. Finished files are the result of years of scientific study combined with the experience of years From aviation: on a flight simulator trained pilots did not notice distractions in the runway, untrained pilots did not anticipate what they would see and noticed the distractions Nurse: I was amazed you got the kid tubed it is so distracting when the grandmother hits the floor ify

7 Change inattention Change inattention / h? Limits of attention Change inattention Limits of attention Narrowing of the choice of diagnostic possibilities too early in the process the correct diagnosis is never seriously considered Recent experience with a disease may inflate the likelihood of it s being diagnosed Example: always comes in threes.. Tendency to seek out data that supports one s original diagnosis, rather than seeking non confirmation data. Tendency toward action rather than inaction. It occurs more frequently in those who are overconfident. Commission bias is less common than omission bias. Spitzer s laws: When you don t know what is going on, call a surgeon. They won t know what is going on either, but they will do something about it.

8 Tendency to characterize a problem in terms of an organ system involved. (Pulm/cardiac) How diagnosticians see things may be influenced by how the problem is framed. Case: 34 week er with failure to pink in the delivery room. Case: it is the middle of the night The tendency to believe that gender is a determining factor in the probability of diagnosis Example: WWB Studies are discrepant.. Unable to find any studies on NNP s, none on neo s Study of ER Physicians (Berk, et al 2008) 829 cases reviewed, 374 errors identified. Physicians with >1.5 years experience were less likely to make an error. Errors not associated with physician age. Inexperience: analytical skills, deliberate. Most of the time this is slow and inefficient (type 2 thinking) Experience: pattern recognition, recalling similar cases. Most of the time this works well.leading to overconfidence. Ignorance more frequently begets confidence than does knowledge. Charles Darwin, 1871

9 Scores on PREP exams (Multiple choice test, encounters with 4 standardized patients, chart stimulated recall) Older physicians performed less well. Positive correlation between experience and diagnostic accuracy. There is evidence that older physicians are particularly influenced by information encountered early in a case. Older adults have a greater tendency to infuse personal experience into problem representations. (a potential explanation is that nonanalytic diagnostic strategies remain strong, but use of analytic confirmation strategies declines) Older adults have more difficulty updating task requirements. Literature suggests that either aging physicians less likely to keep up to date with the medical literature or dependence on Nonanalytic processing makes them less likely to incorporate novel conflicting information Expert skills develop over years The types of errors made by experts are different! How old is the average NNP? If you go to pubmed and search neonate, published in over 512, 789 articles (human neonate 889). How many have you read? Just some observations.. Cognitive error or tool? Framing effect Factors that put us at risk We change our behaviors/orders depending on which neo is on Tend to take on a large cognitive load (busy!) Fatigue Age

10 Factors that may offer some protection Experience We tend to work as part of a group Communication patterns Croskerry Best estimates 5 15% of time (Berner & Graber 2008) Typically have multiple root causes Average of 6 causes per case (Graber et al. 2005) 2/3 system related factors Coordination of care, abnormal result not seen 2/3 cognitive elements Did not gather data appropriately or synthesize appropriately Knowledge deficits rare in mature clinicians (internists..) Any solutions need to include both a system s approach and a cognitive approach On suggested method is to look at Aviation and its use of Situational awareness (Singh, 2006) Situational awareness (perception of situation, comprehension, prediction of consequences), decision making and ability to perform are influenced by: Goals and objectives, expectations Abilities, experience, training System capability, stress and workload Complexity, automation

11 Your brain is always looking for a pattern, once it finds one (real or imaginary) it tends to ignore things that don t fit that pattern There is a limit to the number of things we can attend to Multi tasking myth Affected by fatigue, age, distractions.. Our Brain has both hard wired and learned biases Humans tend to overestimate their abilities 94% of academic doctors rate themselves as performing within the top half of their profession. Doctors have difficulty in recalling errors they make (Norman & Eva, 2010) Develop insight and awareness Consider alternatives Metacognition (reflective approach) Decrease reliance on memory Simulation Cognitive forcing strategies Minimize time pressures Accountability and follow up Feedback Specific training: distinguishing correlation from causation, Bayesian probability Development of expertise Experts perform at highest level of proficiency Speed expertise: Cases, practice and formative feedback (Ericsson, 2008) Conscious reflective thought Be aware of cognitive dispositions to respond Use a checklist Complete history, comprehensive exam, systematic approach, pause and reflect, ensure follow up! Feedback loops High Risk Situation Was the patient handed off to me from a previous shift? Was the diagnosis suggested to me by the patient, nurse or another provider? Did I just accept the first diagnosis that came to mind? Potential Biases Diagnosis momentum, framing Premature closure, framing bias Anchoring, availability, search satisficing, premature closure Called for primary csection at term, report: mom is hepatitis B positive Did I consider other organ systems besides the obvious one? Is this a patient I don t like, or like too much, for some reason? Have I been interrupted or distracted while evaluating this patient? Am I feeling fatigued right now? Am I cognitively overloaded or overextended right now? Anchoring, search satisficing, premature closure Affective bias All biases All biases All biases

12 Called at one hour of age, infant s accucheck 16. The RN has fed the infant, it is 20 min later and the glucose screen is 38. Discovered the mother has some history that was left out. Admitted to NICU orders Infant ate well Weaning? Loose stools Called by lab the lytes are abnormal, should we release them? Called attending orders? Day of life three: infant discharged to WBN to be with mom. Follow up gluc and Na ordered Sodium on DOL 4 was 130 Infant discharged on DOL 4 Case report: What do you know about most maternal medications? An error does not become truth by reason of multiplied propagation, nor does truth become error because nobody sees it. Mohandas Gandhi